Séminaire chercheur invité

Du 30 juin 2026 au 30 juin 2026

De 12h30 à 13h30, en salle du Conseil (1er étage de la faculté de droit, économie et gestion).
Accessible en distanciel sur l'équipe Teams "Séminaires Granem" (réservé aux membres).

Mario Celestino Nardelli est doctorant en « Méthodes quantitatives pour l'évaluation des politiques économiques » (QMPE) à l'Université de Macerata (Italie), sous la direction de Francesca Severini et Rosita Pretaroli.

Ses travaux de thèse portent sur les dépenses de défense et l'analyse multisectorielle en France, à travers la modélisation des tableaux des flux d'approvisionnement et d'utilisation ainsi que des tableaux d'entrées-sorties. Il est accueilli en séjour doctoral au Granem par Pr. Guido Hülsmann, et présentera : 

"Defence Expenditure and Multisectoral Analysis in France"

Abstract:

The French government has committed to spending €413.3 billion on defence over the period 2024–2030. That creates an important structural question: when the state spends a defence euro, where does it end up in the economy: in defence administration, in weapons-systems investment, in domestic output, in imported systems, or in exports? Official national accounts cannot provide a direct answer, because they “hide” defence within civilian aggregates. A defence-resolved supply-use and input-output table was built for France for 2021, based solely on public data that identifies the strictly defined defence industrial base (DTIB) and public defence administration. All splits were reconciled back to their respective published national-accounts totals, producing a higher-resolution version of the official national-accounts table.

The results change the aggregate reading. The strict industrial bundle has a baseline domestic output multiplier of 2.044 and retains 0.708 of each domestic euro as value added; strict military platforms alone activate €2.07 and retain 0.69, against 1.69 and 0.48 in the import-leaky aggregate that contains them. Public defence administration has the opposite profile: an output multiplier of just 1.45, but eighty-nine cents retained at home, because the route is mostly wages. Expressed as demand vectors, the 2024–2030 programming-law routes return between 58 and 89 cents of French value added per procurement euro, with leakage of 11 to 42 cents; the European instruments either map onto these routes, divide by French industrial incidence, or change financing rather than demand. These are structural readings under fixed 2021 coefficients.